metadata
dataset_info:
features:
- name: query
dtype: string
- name: documents
sequence: string
- name: scores
sequence: float64
splits:
- name: train
num_bytes: 5101199648
num_examples: 109810
download_size: 2872383380
dataset_size: 5101199648
configs:
- config_name: default
data_files:
- split: train
path: data/train-*